Are global harvests keeping pace in a warming world?

The world will achieve record food production this year, but with a growing El Niño and future climate change – will it be able to keep up?

It’s been a hot year. The warmest year on record, with several months 1.5°C above pre-industrial temperatures.¹

Next year is likely to be even warmer as the El Niño (the ‘warm’ phase of the natural ENSO cycle) continues into Spring. It could be the first year that we temporarily pass 1.5°C (although probably temporarily).

Food security is my biggest worry about climate change. Many of our current crop varieties could see a decline in yields as the planet continues to warm. In some regions, this could be severe. Now, we’re not helpless in this situation: we can prevent at least some of these losses with different crop varieties, irrigation, farming inputs, and technological innovation.

We’re therefore in a race: can innovation and agricultural productivity keep pace with a changing climate?

So far, the answer has been yes.

Temperatures might have broken records, but so has crop production. I had a look at the latest forecasts from the US Department for Agriculture’s World Agricultural Production reports and database.²

Note that these a preliminary estimates for the 2023/24 marketing year; some of these harvests will continue into 2024, and could still change. But I think are unlikely to move a lot, or change the overall conclusions.

The USDA expects that it will be a record year for many of our crops. I’ve charted some below. These have varying lengths of time series, which is why some are shorter than others.

Rice, corn, soybean, palm oil, cherries, peaches – and several other crops I’ve not included – will see all-time high outputs.

---

Yields of most crops also increased

We shouldn’t just look at crop production alone. We could produce more and more food by expanding farmland, but we don’t want to do that. It leads to deforestation and biodiversity loss.

We want to produce more food with less land, which means higher yields.

While land use for some crops is still rising (this is not surprising – global cropland area continues to expand), yields are also still on the up. This is shown in the chart below for three staple crops: rice, corn, and wheat.

Rice and corn both achieved record yields this year. Wheat was a bit below last year’s high.

---

Some regions have been hit hard by weather events

Of course, there are local disasters beneath the global story.

Some regions have experienced significant drops in crop output because of flooding or drought. As we’ll soon discuss, not every weather event is caused by climate change, but at least some of these impacts will be linked to warmer temperatures.

You might remember the large floods that hit Thessaly in September. This region is a key breadbasket in Greece – its low-lying plains make its soils nutrient-rich and productive, but also prone to flooding. 

While it produces a range of produce including corn, tomatoes, apples, and livestock, cotton was the hardest hit.³ You can see the significant drop in production in the chart below.

That’s just one of many examples: Brazilian wheat had its lowest yield in 6 years due to extreme rainfall; Mexican corn production is down 9% from last year’s record due to drought; wheat in Kazakhstan is down by 27% due to wet conditions.

---

Modern agriculture is much more resilient to extreme conditions

This year the world has experienced a pretty rare El Niño event. These create ‘warm years’, compared to La Niña events that keep temperatures cooler. If you’re interested in what makes this so rare, I’ve left more details in the footnote.⁴ That’s at least part of the reason why this year has been so warm: the underlying driver has, of course, been climate change, but the big anomalous jump we’ve seen is related to other forcings.⁵

We have to go back more than a century to find a good analog of this in the record.

From 1876 to 1878, a rare El Niño, combined with another weather oscillation – the Indian Ocean Dipole – led to severe drought across India, China, South America, and parts of Africa. Tens of millions died from famine.⁶ Millions in India, China, and Brazil.

We won’t have tens of millions dying from famine this time. Yes, there have been severe droughts – you can see it in the large falls in hydropower output in China, and the severe drying in the Amazon. But modern agriculture, global supply chains, and political structures mean we’re much more resilient to extreme events.

Many breadbaskets have achieved and maintained high yields due to agricultural tech and innovations. Yields across the world have doubled, tripled, or more over the last 50 years.

But support is also available for regions where disaster has struck. Aid support can plug acute shortages. Global trade means that imports can fill holes in demand. And in well-governed countries, governments often step in to support affected communities financially.

This wasn’t always the case. Especially not during colonial periods. Look at the history of famines – of which there are many events leading to millions of deaths – and we find very few in democracies. Famines have nearly always occurred in colonies or autocracies.

Global hunger today is still a massive problem: almost 800 million people not getting enough calories. But it’s not because the world can’t produce enough food: it’s caused by large inequalities in production and distribution, poor governance, and exacerbations such as conflict.

---

We need investment into resilient agriculture to continue to out-run climate change

As I said earlier: we’re currently in a race against warming temperatures. Globally, we’ve managed to keep up. But we should be careful about extrapolating this into the future.

Without large investments into more climate-resilient agriculture and productivity, this won’t continue forever.

Optimising crop selections for local weather; developing drought-resistant strains; investing in irrigation and other inputs will all help.

At a larger scale, developing food production processes that don’t rely on outdoor farming at all (such as lab-grown meat, or synthetic fats) could take this to another level. That future might be decades off, but it’s one that we should consider as a changing climate also puts more strain on livestock.

Next year is likely to be a very hot one, with a strengthening El Niño through to April or May. Many analysts already expect a drop in crop output in many countries. Seems like a reasonable assumption. The question is whether we can absorb these shocks at a global level: so far, we’ve done a reasonable job of it.

I’ll be watching the data to see whether we’re still winning the race.

---

1. All sources are in agreement about it being the warmest year on record. Berkeley Earth thinks it could break 1.5°C, while most others think the average temperature will be a bit less – closer to 1.4°C.

The following chart is from the World Meteorological Organization’s State of the Climate 2023 Report.

2. The USDA publishes regular updates of its world agricultural outlooks: these provide summaries of the latest projections of food output and trade. It releases new updates monthly, with data from foreign governments, economic and meteorological data, and satellite imagery.

The data presented here is its latest (December) outlook. While these figures are preliminary – and probably imperfect – I expect they’ll be within a reasonable range of uncertainty. Actual crop outputs in some countries might be out by a few percent, but I doubt the USDA will have somehow missed a significant 5%, 10%, or more drop in crop outputs or yields, especially at a global level.

Note that this data is based on the marketing year “2023/24”, which I’ve shortened to 2023. How the USDA explains marketing years:

“Marketing years refer to the 12-month period at the onset of the main harvest when the crop is marketed (i.e., consumed, traded, or stored). The year first listed begins a country's MY for that commodity (2023/24 starts in 2022); except for summer grains in certain Southern Hemisphere countries where the second year begins the MY (2023/24 starts in 2023).”

3. He, K., Yang, Q., Shen, X., Dimitriou, E., Mentzafou, A., Papadaki, C., ... & Anagnostou, E. N. (2023). Brief communication: Storm Daniel Flood Impact in Greece 2023: Mapping Crop and Livestock Exposure from SAR. Natural Hazards and Earth System Sciences Discussions, 2023, 1-16.

4. The ENSO cycle oscillates between a year or two of El Niño (warm) and La Niña (cool) phase, with a neutral year (or two) between. 

This year the world came out of a ‘triple-dip’ La Niña event, where we’ve been in the cool phase of the cycle for three consecutive years. During La Niña, heat tends to accumulate in the ocean surface waters – keeping atmospheric temperatures slightly cooler. That means that after a triple-dip event, there is a lot of heat in the sea surface, waiting to come out.

Now, we’ve had ‘triple-dip’ events before: mostly recently from 1973 to 1976, and from 1998 to 2001.

But what has also been rare is that they tend to follow very strong El Niño events and the world flipped very quickly from the La Niña, straight into an El Niño earlier this year. Usually, there is a longer ‘neutral’ period between these phases, where heat will be removed from the surface ocean more slowly. Without this more moderate period, heat has very quickly moved from the surface waters to the atmosphere, resulting in much more rapid warming.

This recent event is challenging scientists’ understanding of the variability in ENSO cycles.

5. There has been lots of debate about what has caused a big jump in temperatures this year. The underlying driver is greenhouse gas emissions. But this is not the cause of the anomalous rise.

Hypotheses include the Hunga Tonga-Hunga Ha’apai volcanic eruption, solar cycles, the impact of removing aerosols as we clean up our air, and the El Niño.

I don’t have a horse in this race – and it’s not my area of expertise – so will leave it to the proper climate modellers to fight it out.

From what I’ve seen from climate experts, the individual impact of the volcanic eruption, solar cycles, and aerosols are likely to be relatively small. The primary driver, then, is probably El Niño (and perhaps internal climate variability, which is difficult to quantify precisely).

6. There are a range of estimates for the total death toll of this famine event. Most are in the order of tens of millions. My colleagues, Joe Hasell and Max Roser, have an estimate of 20 million in our large historical dataset on famines. However, not all countries are included due to poor data availability so it could be higher. Upper estimates can be as high as 50 million.